Using IR Lasers Instead of Fiber

Artifice_Eternity writes: "Can't deal with the trouble, time or expense of digging up the street to get fiberoptic cable to your building in the big city? There's another way...infrared line-of-sight infrared lasers between your building and another one nearby. Repeaters and redundancy can keep the chain going reliably for miles, with gigabit data transmission rates."

Comments

Weather

[CAIMLAS]

What about cloudy days, storms, or even ice on the windows (provided the instrument is housed)? Would these things effect bandwidth? BOFH answer #556454: You can't get your files off the server right now because the cloud cover is too thick.

infrared?

[khuber]

"infrared line-of-sight infrared lasers" All products are approved by our Department of Redundancy Department. -Kevin

Plywood Fade :-)

[billstewart]

A number of years ago, a friend of mine was at University of Colorado. They had two computer centers which were connected by an infrared laser which was pointed out the window of one building to the other building. It had minor data loss during snowstorms, but was pretty reliable. One day the window broke, and a repair person came by and put a piece of plywood over the window until they could get a piece of glass big enough to repair it properly. He didn't understand why all the computer people started yelling at him.... after all, *he* couldn't see the invisible light beam going through the broken window :-)

READ THE ARTICLE, PLEASE

[Artifice_Eternity]

The "repeaters" in this case can be placed indoors, in front of a window. One of the reasons for developing this system was to bypass the trouble and expense of rooftop transmitters.

And note that even in my summary I mentioned redundancy -- multiple IR beams are designed to compensate for bad atmospheric conditions -- and each hop in the network is a short distance for the same reason.

The good, the bad.

[Restil]

This is good for many reasons. The cost is
primarily in the one time purchase of the equipment. And it makes a lot of sense where line of sight is a shorter distance than a fibre cable must travel (tops of buildings in a congested city). These dont interfere with radio freqencies, so you don't have to concern yourself with noise, or creating noise. And its unlikely a backhoe will ever be a problem (as long as it doesn't block the line of sight).

The downside is the line of sight. You ALWAYS have to have line of sight. Rain, fog, clouds, trees, idiots with signs, they all can cause problems. Short distances are less of a concern, but you still have to maintain an almost perfect orientation. A little gust of wind can have you dropping packets.

But its probably a better solution than fibre where running fibre isn't an economically feasable solution. But no matter how good this is, fibre has far greater potential capacity, even though we don't yet have the technology to use all of it. It doesn't make sense to start building the internet backbone out of these things.

-Restil

It's been done for over a decade. "Arclight"

[Ungrounded Lightning]

Infrared line-of-sight links have been done for well over a decade. Datapoint's Arcnet had an infrared link device called Arclight that they used for a line-of-sight link for several miles. It would flake out in rain or fog - because rain and fog both absorb and refract infrared. So it might as well have been as opaque as black smoke.

Arcnet was a self-healing token ring network with an underlying broadcast topology. So if two buildings were connected by Arclight and it went down, the network split into two rings, and when it came back up it healed into a single ring. Reconfig took miliseconds so it was no big deal.

You may not have heard of Datapoint. But have probably heard of the Intel. Seems Datapoint had a discrete-component standalone computer/smart terminal which was the basic node in their network - a diskless-workstation, fileserver, compute-server archetecture. They cut a deal with a semiconductor company called Intel to try to port their instruction set to a silicon chip for the next generation. But the resulting chip was too slow, so they went with another discrete component solution.

And Intel had cut the deal so they could sell the chip. So they took the chip to market, perhaps with a few tweaks, as the 8008 - first in the line that continued with the 8080, 8086, 80x86, Pentium, ...

Once again ... the pitfalls ...

[pgrote]

Yes, as many have commented this is old news. There haven't been any breakthroughs on this in several years. Let me rephrase that ... no significant breakthroughs.

Once a year we have a customer come to us to ask about this option. Once a year we trot out our proposal, rerun the numbers and submit it. It always comes down to two things:

1) Speed
2) Reliability

The speed of the lines is fantastic when you have a clear day and relative distance is kept. Any atmoshperic conditions out of the ordinary will kill the line. Now, if you're only interested in using it for a mail gateway or to transfer data for a nightly batch cycle it rocks. For regular WAN access you'll be answering the phone from the folks on the floor.

Reliability is a concern past the speed. Keeping the connections is sometimes more an art than science. The article does have an interesting take on parallel transmissions, but commerically available products are cost prohibitive to make it a viable alternative for most businesses.

The best line from the article is: "Transmitting lasers through the air and modulating them at gigabit rates is a new and potentially powerful development," said Dr. Daniel Leslie, a physicist at Trex Enterprises who is familiar with the new systems.

Now, if you want to see something cool check out Dr. Leslie's company web page: http://www.trexenterprises.com/laserrad.html

If only we could work a Pringles can into this ...

lots of applications, lots of promise

[hyrdra]

Hmm, this is nothing new. You, yourself, can buy a several watt (yes watt) IR diode and modulate it using an AOM to at least 100's of Mbits a second and more very easily for under $1500. This would easily reach a target for miles if the reciver made use of a dichroic narrowband filter and some good ECC. However, as mentioned, line of sight must be maintained.

Divergence, not mentioned in the article, is also an issue. Especially with laser diodes, it doesn't matter what kind of miracle anamorphic lens system you have to decrease the divergence of the beam, becuase pretty soon that pencil thin dot is going to become several feet in diameter. THIS is what accounts to loss more than so called 'atmosphere' causes. Photodiodes/transistors operate at a power/cm^2 ratio, and the lower this is out of the rated area the more noise. So when the beam spreads out, the concentration of power thins out and you get noise because although all your signal is getting there, you can only sample a small fraction of its power. Having low divergence also works against you because it makes the system much more difficult to align. I would start with a very large beam, just enough to get a signal, and then progress to the smallest beam possible. Vibrations at the transmitter site will likely limit this, as tiny shifts in movement only a mm will cause the beam to jump several feet miles away.

Many here have mentioned the speed is on the slow side for this technology. Well, folks, this is optics and that means you can do things in parallel. If you need more speed, just shift the wavelength of the diode and multiplex it in. This is the same principle behind DWDM systems, only it's in freespace. You don't even need a fancy FB diode to do it -- most commercial diodes have a 30 nm linewidth, and by controlling the voltage and temperature you can easily shift up or down. In any case, adding another same-wavelength line is just as easy as adding another transmitter/receiver pair at either end. If only you could do that for fiber. Instead, you have to dig up the streets.

I have had the pleasure of working with a system from Coherent that really makes free space communications shine. The system automatically adjusts and aligns itself via electronic gyros and GPS. It tells you if the current location even has any type of line of sight and if it does it zeos in on the beam (e.g. "I'm pointing S-SW, can you see me?"). The hardest thing is you must have a current connection to the other end while performing the alignment, but this was easily accomidated for at my location with a cellular modem.

This stuff is really cool and there are definate applications for anywhere that has good line of sight. For example, cell towers frequently have good line of sight to one another, so this technology would make sense for that application. There are enough towers that the network could be constructed in a serial or star configuration, without the need for many land lines near the tower.

What would be even cooler would be somehow using the high voltage transmission towers and installing a small, low cost module on each one to jump from tower to tower, or even pole to pole. Since it's optical, you don't have to worry about interference or expensive shielding (yes, there are all-optical transceivers out there).

These are just some ideas but the technology itself finally seems to be maturing. There are lots of current applications and it seems that although most carriers have loads of dark fiber underground, so the cable isn't really the problem, but maybe these companies will help drive bandwidth prices down by enabling small yet very fast ISPs to pop up and use the technology without having to haggle over ground cable. The Internet Revolution per se, isn't going to continue until we all have true broadband (10 Mbps or more, preferably 100 Mbps) to the curb for $19.95 a month.